Post-quantum cryptography in banking: why start now

There is a threat that banking cannot afford to ignore, even if it still seems distant. The day a quantum computer is capable of breaking the cryptography that today protects our communications and transactions. Post-quantum cryptography is the answer to that scenario, and the reason why serious institutions have already started to move.

What post-quantum cryptography is

Post-quantum cryptography is the set of algorithms designed to withstand attacks from both classical and quantum computers. The systems we use today, based on mathematical problems that a powerful quantum computer could solve quickly, would be compromised. The new algorithms rely on problems that remain difficult even for quantum computing.

This is not science fiction: standardisation bodies have already selected the first post-quantum algorithms, and migration has stopped being a theoretical debate to become a real project.

Why migrate now and not later

The “harvest now, decrypt later” threat

The most urgent risk is that an attacker captures encrypted data today to decrypt it in the future, when quantum power allows. For banking, with data that must remain confidential for decades, this threat is real from now, not ten years away.

Migration is slow

Changing the cryptography of a complex organisation takes years. You have to inventory where it is used, prioritise and migrate without breaking anything. Starting late is not an option, because the process cannot be sped up at will the day the threat materialises.

How to plan the transition

I recommend starting with a cryptographic inventory: knowing where and how encryption is used throughout the organisation. From there, prioritise the systems that protect long-life data and move towards agile architectures that make it possible to change algorithm without redoing everything, which planning fits with cybersecurity risk management and with the resilience requirements of the DORA regulation.

Why the urgency is today

The reason I insist so much on starting today is a risk that many executives do not yet have on their radar. The “harvest now, decrypt later” strategy. An attacker can intercept and store encrypted communications that are undecipherable today, waiting for a sufficiently powerful quantum computer to allow them to open them in a few years’ time. For data with a long life cycle — financial records, contracts, medical information — that threat is real from this very moment, even though the quantum computer does not yet exist.

This completely changes the urgency of the problem. It is not about preparing for a future risk, but about protecting today information that will still be sensitive when current cryptography no longer holds. That is why, when an organisation tells me it has plenty of time, I ask a simple question. How much of your current data will still matter in ten years? The answer almost always dissolves the false sense of calm.

Cryptographic inventory: the first real step

Before migrating anything, it is worth knowing what you have. The most underrated step, and at the same time the most useful, is to draw up a complete cryptographic inventory. Which algorithms, which key lengths and which certificates are used in each system, library and provider. Without that map, any migration plan advances blindly. With it, you can prioritise by criticality and tackle first the systems that protect the most sensitive and longest-lived data.

This migration is not an isolated project, but part of a broader security architecture. It fits naturally into a strategy of Zero Trust in banking and regulated environments, where trust is never assumed and each layer of protection is reinforced independently. Thinking of post-quantum cryptography within that framework avoids treating it as a patch and turns it into another pillar of long-term resilience.

Crypto-agility: designing for change

The ultimate goal of the migration is not simply to replace one algorithm with another, but to gain crypto-agility. The ability to change algorithm with the least possible effort when the standard evolves. Post-quantum algorithms are still maturing, and it is likely that there will be adjustments in the coming years. That is why I recommend designing systems so that cryptography is a decoupled and replaceable piece, not something embedded in the business logic.

In practice, this means abstracting cryptographic functions behind clear interfaces and avoiding rigid dependencies on a specific library. A crypto-agile organisation will be able to adopt the next generation of algorithms with reasonable effort, while another with cryptography hard-wired into every application will have to repeat a costly project each time. Investing today in that flexibility is, in my opinion, as important as choosing the first replacement algorithm well.

Conclusion: preparing before the storm

Post-quantum cryptography is a textbook case of long-term risk management. A threat that is unlikely in the short term but catastrophic in impact, which demands acting before it happens. The bank that starts migrating now will arrive prepared. The one that waits to see the problem will arrive late to a process that does not allow for haste.

Frequently asked questions about post-quantum cryptography